What do you like best about TIMi?
In my 22 years in the profession and ten working with TIMi, I could write a book about all the true value TIMi has brought to me & Volvo Cars compared to the competition, and I have used and tested many of them.
So if you don't have it accessible in your toolbox yet, download the free version of Community Anatella (ETL) at once and start to learn. You will thank me later.
- When I now think back on my years working with TIMi, it never disappointed me, and that is a statement by itself as you sooner or later always get in trouble in your project. TIMi support was there in a few short hours to assist. I even got custom functions/actions coded in a very short time frame. Try this with SAS, IBM, SAP or other giants. What was done in a few hours would take weeks unless you are paying at least 10x more for instant support. But let's get to the details and examples.
So what is it to like?
If you work with big data, Anatella will be your closest friend and companion to support your corporate tools, whatever you use. I worked with Teradata, Aster (big data tool), Spark, and more. So let's take a simple real-life example. We have hundreds of applications generating terabytes of data from production to marketing. To get any value out of your analysis, you will need to join tables, and your company bought a lot of in-house capacity with, say, IBM, in our case Teradata. Whatever you buy, it will not be enough. It will get filled quickly by consultants writing inefficient union and sortby statements in SQL running daily for different applications, both critical and non-critical. The point is it's breathtakingly expensive to buy a license for additional capacity. Unless your company is in the business of printing money, it can take months to years before additional capacity will get approved. So your queries will be waiting for days/weeks until you get the data you need. And if you have tables with gigabytes/terabytes of data to be joined, sorted and, filtered, you may be out of luck. It eats too much of the precious capacity that is needed for critical production apps. BUT select statements are quick and resource-efficient. You ask your IT to dump tables you need to your local machine with 4 to 8TB NVMe disc, and from there, you can process with Anatella to do the joining and ETL. It's very, very often faster in Anatella, but also, you have better control over the process of data transformation, so you know what you get. This is getting long. You get the point. This was specific to my company & my working situation, and your reality is different. Still, you will for sure get in a position where TIMi/Anatella is something you could not live without.
In short, the Pros:
- you can shuffle TBs; sometimes, the only way to prepare your ADS (analytical data set) is Anatella. Don't believe that if you have all the data in one place like SAP, you can do it all. The toolbox on big platforms is always limited
- robust, don't crash like rapid miner or python when your RAM gets filed. If you use R, you are toast, and Spark has its problems.
- flexible, most if not all database, cloud connection as well as R, Python integration where you can handle the heavy stuff with Anatella and pass small and ready ADS to R or Python. It works. I have done amazing things with this approach.
- surprisingly fast and memory-efficient. If you have a decent desktop or laptop, you can do a lot. Change to 8TB NVMe memory, and you can do it all.
- easy to follow step by step to track for errors. These are often more difficult to spot if you code in python or R.
- easy to present and demonstrate to management to show progress and what can be done as proof of concept or pilot before taking it to production (try to show python or SQL code in the meeting with higher management, and watch the reaction)
- build stand-alone tools/applications where data structure frequently changes, like surveys. Junior analysts can then maintain these applications.
- you can do a lot with the free community version. Think about all the efficiency you could get if you deployed Anatella to all analysts in the company. I did it for the 6-Sigma community as they sometimes use big data sets. 6-Sigma statistical toolbox, Minitab can't handle it.
About the TIMi Automated Modeling
- very fast on big data sets compared to other software, I have run 100's of models for each of my "difficult to solve" root cause analysis projects, where big money was at stake in warranty costs. In this kind of projects, you need to tweak your inputs a lot as the root causes are not necessarily the variables with the highest importance in your model, but the combination of the variables is the key. Python or R would just die on me if I used this amount of data. And also, many competitors would never finish the job because of time. TIMi core has been optimized for running this kind of task.
- In predictive analytics, TIMi has given me significantly better results if discrete variables were involved. I used TIMi as one of the many tools in my toolbox, as I have other favorites, and each of these tools has its place.
- One essential part of using TIMi for me is access to the experience of the TIMi support team. Since they have already done many impressive projects, they often have some hints to help on the way if needed. The reason that the software is so excellent is that these guys have their hands deep in real-life customer problems and know what is required in order to get the project done. Review collected by and hosted on G2.com.
